If there is one principle of criminal justice systems that stands deeply rooted in all countries, it is that of fair trials. And a fair trial, especially for refugees entering the West from Syria and Afghanistan, demands a fair process of interpreting from their native language to English. It is in this context that the seemingly simple legal role accorded to an interpreter in a number of jurisdictions—to be an arbitrary, impassive, neutral actor who merely acts as a machine, as a mechanical mouthpiece of the defendant, objectively converting the speech from one language to another becomes a fiction of law than one of the ground realities.

The Inherent Deficiencies of Translation

Translation as a Mechanical Process

Most legal systems also adopt a view of translation as a solely mechanical process requiring verbatim literal translation without any changes whatsoever, when it in fact is a highly normative process which requires interpreters to change words and phrases when translating.

Especially given that refugees come from countries with vastly different cultures and contexts, phrases in the source language that are fairly clear might have multiple meanings when translated to the target language or vice versa. For instance, in a trial involving Spanish migrants, the word “camioneta” was translated to a “station wagon or a small truck”, and “gafas” was interpreted as “glasses or sunglasses”. Such a seemingly minor change in interpretation of gafas as glasses in one part of the translation and sunglasses in another leads to lower credibility when such interpretations are perceived as contradictory statements. It also makes it sound like the defendant is less sure of whether it was a station wagon or a small truck, when in fact the original speaker was quite clear in his conception of the object. Asylum or racism as a word does not exist in explicit terms in Mam. This, therefore, requires interpreters to supplement their translations with helping words, which might not be strictly in accordance with guidelines prescribed requiring verbatim translation but are nevertheless necessary.

The cultural connotation attached to a word is also significant. In the trial of a father, fleeing persecution from FARC, the defendant had received death threats from the FARC for being an informant, through a dead toad left on his car as a representation of his family’s fate in Columbia. The interpreter translates the word “Sapa” (meaning toad) using the English word “frog” which just does not capture the connotation attached to the word “sapa” in Spanish. An equivalent in English would have been the word “rat”. In such circumstances, a verbatim translation does not capture the meaning of the original source.

In a similar case as the above, the defendant attempts to explain that his fear of the FARC guerillas meant that the family had to leave the village unnoticed. “De una maner diplomatoca” is a figure of speech he uses, with the understanding that they could not leave the village in an overt manner. The interpreter, providing a literal translation, interprets this as “We were gonna be diplomats about it….”. This is immediately succeeded by the attorney’s question “Although you feared for your family, you didn’t want to leave the village because you wanted to be diplomatic, Sir?” casting significant doubt on the defendant’s credibility about the urgency and terror the family faced in Columbia, and mis-interpreting the underlying meaning the defendant wanted to convey, which was the caution with which the family had to flee because their lives were at stake.

Note that in a lot of the circumstances, a clarification from the original speaker themselves about the words or explanations of context would have made the process easier. But, in a context where interpreters are reprimanded for clarifying the translation to the prosecutor and ignored when they draw attention to potential deficiencies in their translation— it is much harder to have those clarifications.  

Ultimately, it is not humanely possible, and in several cases, misleading to have a verbatim word-for-word translation of testimonies. A 2011 study conducted in the United States corroborates this—the majority of people in the study deviated from the client’s original utterance, and none of them consistently produced translations that were not one hundred percent accurate. Note, that the right to equality is not about having the same kind of trial procedures, but it is about asylum seekers being heard to the same extent. In the race to pretend like the bilingual courtroom is just the same as a monolingual setting with a mere substitution of words from one language to another, courtrooms rarely acknowledge, let alone try to correct for the misleading statements that interpreters say in the name of verbatim translations.

Translating tonal variations and “inconsequential” phrases

An interpreter also has the capacity to make the witness testimony sound more antagonistic than it really is, stronger and sure-sounding than intended, and conversely, less challenging than the original, all by slight tonal variations, changes in phrases used, and what would be considered minor additions or omissions. Interpreters have the power to make testimonies cast more (or less) culpability and trustworthiness, impact the juror’s evaluation of the defendant’s intelligence and competence, and hence play a significant role in both the determination of guilt and sentencing.[1] A simple change in the article used, from “Did you see a glass” to “did you see the glass” increases the assumption that a glass does exist, increasing the likelihood of a witness recalling to have seen the said object, even if they have actually not. A change in structure from a “what” to a “did you” question can affect whether the witness presents a lengthy explanation or a short Yes-No answer respectively. 

An analysis of the interpreter’s translation in a trial involving a woman from Columbia, seeking asylum in the United States with her son elucidates how minor changes in the words used can have major impacts by weakening testimonies. Here the defendant explains that she was involved in juvenile brigades against the FARC, perhaps to emphasize her active, military participation and thus strengthen her argument that she would be persecuted by the FARC if deported. But the interpreter’s translation of “And I got, uh, involved in juvenile groups” gives a significantly softer understanding of the defendant, diminishing the force of the defendant’s word choice.  This becomes even more significant considering that the defense attorney picks up the words of the interpreter “groups” when further asking questions, with these becoming the phrases which characterize the defendant’s experiences. Ultimately, in this trial, the defendant was not able to convince the judges that she had a credible fear of persecution.

The subconscious omissions and additions of “uhs” and “probably” further determine the extent of hesitation that can be inferred from a testimony. Interpreters often ignore (and understandably so) what seem to be unimportant elements of speech like “ah/uh” “este/uhm” or “pues/well” of the defendants. At the same time, they frequently use “uhs” which are often their own and not a direct translation, which results from the mental strain from the process of interpretation.

There are also conscious, active changes to questions that can have detrimental impacts. A change in interpretation in an Austrian trial from the judge’s question on the “threatening behavior and attempt to hit the policemen” to “you took a swing, you wanted to hit the policeman” has completely different connotations attached. The latter is far more accusatory in tone and includes another charge of an intention to hit the policeman, leading the defendant to defend himself against such an accusation rather than answer the judge’s question.

This is precisely why a right to fair linguistic interpretation is critical. If it is true that even a minor change can affect the outcome of a trial massively, then one can only imagine the extent to which major changes due to lack of training will affect trial outcomes.

A Catch-22

Translating verbatim from one language to another is neither humanely possible nor desirable given there are a number of words that exist in one language that do not have a clear equivalent in another. But this presents a challenge in itself— by not translating verbatim, the interpreter by choosing which parts of the speech to translate, which words to emphasise, which spaces to pause, is subconsciously diminishing the original speaker’s command over what they would like to present.

But in order to come as close to a verbatim translation as possible, speakers are asked to give shorter answers to questions, are interrupted in between which disrupts their continuity[2], and are discouraged from speaking, which in turn impairs their ability to present a fluent, detailed explanation that builds credibility and ensures that their story is heard in all its entirety.

Furthermore, it is not enough to translate what is being said to or by the defendant. The defendant must understand the court proceedings as a whole, the judge’s comments, the prosecution’s arguments, and objections, the attorney’s claims, etc. According to one study of England and Wales, in around 40 percent of the cases, interpreters remained silent for long periods of time, not translating any of what was being said to defendants or witnesses. This is further entrenched due to the complex nature of legalese and attempts by judges to silence such translation under the reasoning that it is disruptive to court proceedings. Equally liable were lawyers and practitioners, who frequently pretended that the court was a monolingual setting, giving no pauses for the interpreter to translate what had been happening.

These issues are further exacerbated when the judge who is responsible for ensuring that the proceedings are conducted in a fair manner, often is not in a position to determine when the interpreter has not faithfully conveyed the defendant’s views. In the district court of Denmark, when conflict arises in the courtroom from the audience due to erroneous interpretation of French, the judge who was not aware of the linguistic errors, continued the proceeding with the same interpreter.[3] In the majority of cases, no one in the courtroom understands the language spoken to create accountability on the interpreter. Secondly, while defendants and witnesses might testify in foreign languages, the interpreter’s translation is the only one which stays on the written transcript without a mention of controversies over crucial words translated. This leads to a situation where the problem can neither be corrected as it happens nor identified through retrospection.

Throughout criminal justice systems, we already recognize the right to record the judgement, record the testimony and record police investigation procedures. It is therefore ironic that where it is most important to record the different versions being translated, the rights of asylum seekers to have that record has gone unnoticed.


A first step and a suggestion would be to record the defendant’s original testimony on the written record which both helps create accountability to the interpreter and also aids in providing further research on potential solutions.

Secondly, allowing interpreters to seek clarifications after taking permission from the judge to do the same, to elucidate the different meanings a sentence could potentially have, to explain why a certain language might not have the necessary words to answer a question is critical to having a more effective translation system.

A third step that is suggested is to ensure that there are two interpreters in a courtroom—one who stays by the side of the defendant to translate what is happening in the court proceedings, and another who can translate witnesses’ and defendant’s testimony. This ensures accountability but also helps to correct misinterpretation, and raise objections when relevant.

Note, it is broadly recognized that asylum seekers have a right to a trial in the language they know, rather than an alien language they are unaware of. But that right is meaningless if it is not backed by an equally substantive claim to an interpretation that is accurate.  Asylum seekers are already in a vulnerable position, often having crossed the most dangerous paths to seek some protection. To deny them refuge without a trial would be a gross violation of due process, and it is unclear why a policy that bars us from hearing them is any different.

[1] Susan Berk-Seligson, The Bilingual Courtroom: Court Interpreters in the Judicial Process 7-10 (The University of Chicago Press, 2d ed. 2017).

[2] Miriam Shlesinger & Franz Pochhacker, Doing Justice to Court Interpreting 85-99 (John Benjamins Publishing 2010).

[3] Shlesinger, supra note 2.

This article is authored by Saranya Ravindran, a second-year student at the NALSAR University of Law, Hyderabad.

Categories: ArticlePosts

Notify of

Inline Feedbacks
View all comments
Would love your thoughts, please comment.x